Potentiometric Selectivities of Ionophore-Doped ... - ACS Publications

4 Jan 2019 - Well-accepted models predict how these selectivities can be optimized by ...... Bühlmann, P.; Chen, L. D., Ion-Selective Electrodes With ...
0 downloads 0 Views 659KB Size
Subscriber access provided by TULANE UNIVERSITY

Article

Potentiometric Selectivities of Ionophore-Doped IonSelective Membranes: Concurrent Presence of Primary Ion or Interfering Ion Complexes of Multiple Stoichiometries Ibrahim Yilmaz, Li D Chen, Xin V. Chen, Evan L. Anderson, Rosenildo Correa Da Costa, John Gladysz, and Philippe Bühlmann Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b05196 • Publication Date (Web): 04 Jan 2019 Downloaded from http://pubs.acs.org on January 10, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Potentiometric Selectivities of Ionophore-Doped Ion-Selective Membranes: Concurrent Presence of Primary Ion or Interfering Ion Complexes of Multiple Stoichiometries

Ibrahim Yilmaz,a,b Li D. Chen,b Xin V. Chen,b Evan L. Anderson,b Rosenildo Correa da Costa,c John A. Gladysz,c and Philippe Bühlmann*,b

a b

Kamil Ozdag Science Faculty, Karamanoglu Mehmetbey University, 70100 Karaman, Turkey

University of Minnesota, Department of Chemistry 207 Pleasant St. SE, Minneapolis, Minnesota 55455

c

Department of Chemistry, Texas A&M University, P.O. Box 30012, College Station, TX 77842

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

Abstract The selectivities of ionophore-doped ion-selective electrode (ISE) membranes are controlled by the stability and stoichiometry of the complexes between the ionophore, L, and the target and interfering ions (Izi and Jzj, respectively). Well-accepted models predict how these selectivities can be optimized by selection of ideal ionophore-to-ionic site ratios, considering complex stoichiometries and ion charges. These models were developed for systems in which the target and interfering ions each form complexes of only one stoichiometry. However, for a few ISEs, the concurrent presence of two primary ion complexes of different stoichiometries, such as ILzi and IL2zi, was reported. Indeed, similar systems were probably often overlooked and are, in fact, more common than the exclusive formation of complexes of higher stoichiometry unless the ionophore is used in excess. Importantly, misinterpreted stoichiometries misguide the design of new ionophores and are likely to result in the formulation of ISE membranes with inferior selectivities. We show here that the presence of two or more complexes of different stoichiometries for a given ion may be inferred experimentally from careful interpretation of the potentiometric selectivities as a function of the ionophore-to-ionic site ratio or from calculations of complex concentrations using experimentally determined complex stabilities. Concurrent formation of JLzj and JL2zj complexes of an interfering ion is shown here to shift the ionophore-to-ionic site ratio that provides the highest selectivities. Formation of ILn-1zi and ILnzi complexes of a primary ion is less of a concern because an optimized membrane typically contains an excess of ionophore, but lower than expected selectivities may be observed if the stepwise complex formation constant, KILn, is not sufficiently large and the ionophore-to-ionic site ratio does not markedly exceed n.

Keywords: Ion-selective electrodes, Ionophore, Ionic Sites, Complex Stoichiometry, Potentiometric Selectivity, Fluorous Membranes

2

ACS Paragon Plus Environment

Page 3 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

When ionophore-based ion-selective electrodes (ISEs)

1-8

were first introduced, their

membranes were doped with ionophores, but no deliberate effort was made to introduce ion exchanger sites. Only later was it shown that ISE membranes with electrically neutral ionophores but no added ionic sites gave the well-known Nernstian responses to the target ions only because the matrix polymer and plasticizers contained charged impurities that functioned as ion exchanger sites. 9-13 This led to the deliberate introduction of highly hydrophobic ions to provide for ion exchanger sites. Moreover, it was shown that the ratio of the ionophore and the ionic sites in an ISE membrane can change the potentiometric selectivity by many orders of magnitude.

14-18

This is true both for neutral and

electrically charged ionophores. The reason for the large effect of the ionophore-to-ionic site ratio on the selectivity lies in the stoichiometry of the target ion complexes in the ISE membrane.4, 7, 19 Because of the requirement for electroneutrality in bulk phases, the bulk of an ISE membrane must contain an amount of exchangeable ions such that their total charge equals (but is opposite in sign to) the total charge of the ion exchanger sites (and the ionophore, if the latter is electrically charged). If the ionic site concentration is chosen to be too high, the total concentration of target ions in the bulk of the ISE membrane is too high as well, and there is an insufficient amount of ionophore available for the formation of complexes with the target ion. On the other hand, in a highly selective ISE membrane, all but a minuscule amount of the target ions is bound in the form of ionophore complexes, and the membrane contains an appreciable excess of free ionophore. In most of the published work on ionophore-based ISEs, the ionophore has been assumed to form only complexes of one stoichiometry with the target ion, although different stoichiometries for primary and interfering ions were considered thoroughly.14-18 This led, e.g., to the recommendation that membranes doped with an electrically neutral ionophore that forms 1:1 complexes with the monovalent target cation I+ but 1:2 complexes with the monovalent interfering cation J+ should be doped with 71 mol % anionic sites to achieve optimum potentiometric selectivity.18 3

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 29

Only a small number of publications mentioned the possibility of multiple complex stoichiometries for a given type of ion under different conditions (such as in membranes with different ionophore-to-ionic site ratios).20-23 Moreover, there are few examples of the concurrent presence of target ion complexes of multiple stoichiometries in ISE membranes. Probably the first example for such a system was given by the Mg2+ response of a Mg2+ ISE in a background of Ca2+. It was speculated that this Mg2+ response was affected by the concurrent presence of 1:1 and 2:1 complexes of the ionophore with Mg2+, but a quantitative discussion was not provided because complex formation constants were not available at the time.20 More than ten years later, a report on [9]mercuracarborand-3 as chloride ionophore gave quantitative evidence for 1:1 and 2:1 complex formation with the target ion Cl-.21-22 It was understood that these two types of complexes coexisted in the sensing membrane, but the effect of the ionophore-to-ionic site ratio on the potentiometric selectivity was only referred to in the context of lower detection limits. A more comprehensive discussion of primary ions that formed complexes of multiple stoichiometries was provided for fluorous24-25 membrane ISEs doped with an electrically neutral fluorophilic crown ether ionophore. In that case, K+, Cs+, and NH4+ each formed 1:1 and 2:1 complexes, and it was shown how the concentration of these complexes in the sensing membrane depends on the sample composition.23 For example, when membranes doped with this ionophore and 71 mol % anionic sites were exposed to samples that contained K+ as the only cation, this resulted in formation of substantial concentrations of 1:1 and 2:1 complexes of the ionophore and K+ in the sensing membrane, while the concentrations of both the free ligand and free K+ in the ISE membrane remained very low. This thermodynamic model explained the apparently super-Nernstian responses26-28 of those electrodes to K+ when measured in a background of an interfering ion.23 Interestingly, the closely related technique of ion-transfer voltammetry has been used to determine sequential binding constants of ions to ionophores.29-32 However, since ion-transfer voltammetry is based on the use of hydrophobic sensing membranes that comprise an ionophore and a hydrophobic

4

ACS Paragon Plus Environment

Page 5 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

electrolyte but no ion exchanger sites, complexes with higher stoichiometry are more likely to be formed preferentially unless the ionophore is depleted or complex stabilities are particularly small.33 Notably, in the above mentioned case of [9]mercuracarborand-3,21-22 the concentration of free ionophore was not considered quantitatively, and, in the case of the fluorophilic crown ether,23 the free ionophore concentration was low throughout the whole range of the calibration curve. Moreover, in none of the above cases was an attempt made to predict (i) the concentrations of the different complexes and free ionophore in a wider range of ionophore-to-site ratios and (ii) the resulting effects on potentiometric selectivities. To this end, we briefly described earlier the perhaps counterintuitive finding that, for low ionic site concentrations, membranes doped with an ionophore that forms multiple n:1 complexes with a monovalent ion will have lower potentiometric selectivities than membranes doped with an ionophore that only forms 1:1 complexes.7 We are reporting here the theory to explain this finding and discuss the reasons that explain how the ionophore-to-ionic site ratio affects potentiometric selectivities in systems with multiple stoichiometries, using as an example a hypothetical ionophore that forms complexes of 1:1, 2:1, 3:1, and 4:1 stoichiometry. This is followed by experimental data for a fluorophilic Ag+ ionophore, which appears to be the first example of a highly selective ISE membrane that contains 2:1 and 1:1 complexes in the presence of a substantial excess of free ionophore. Finally, the effect of multiple stoichiometries on the optimum ratio of ionophore and ionic sites is discussed. Both the theoretical and the experimental results illustrate (i) that ignoring higher complex stoichiometries may result in the use of suboptimal ionic site concentrations and, therefore, poor selectivities, and (ii) that even a detailed experimental plot of selectivity versus the ionophore-to-ionic site ratio does not always readily reveal the concurrent formation of multiple complex stoichiometries.

5

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

EXPERIMENTAL SECTION Reagents. All commercial chemicals were of high purity and were used as received. Perfluoroperhydrophenanthrene and tetraethylammonium acetate were purchased from Alfa Aesar (Ward Hill, MA) and GFS Chemicals (Powell, OH), respectively. Sodium tetrakis[3,5bis(perfluorohexyl)phenyl]borate, 1,34-35 and 1,3-bis(perfluorooctylethylthiomethyl)benzene, 2,36-37 were synthesized as reported previously. All sample solutions were prepared with deionized and charcoal-treated water (18.2 MΩ cm specific resistance) purified with a Milli-Q PLUS reagent-grade water system (Millipore, Bedford, MA). Rf6

Rf6

Rf6

Rf6 Na

B Rf6

Rf6 Rf6

S

Rf8

Rf6 1

S

Rfn=(CF2)n-1CF3

Rf8 2

Figure 1. Structure formulas of the fluorophilic site, 1, and the Ag+ ionophore, 2. Ion-Selective Membranes. Fluorous stock solutions were prepared to contain 1.0 mM ionic site (1) and 0 or 5.0 mM ionophore (2) in perfluoroperhydrophenanthrene as the solvent, and were stirred for at least 24 h to ensure complete dissolution. Sensing phases that contained ionic sites (1.0 mM) and ionophore (in the concentration range from 0.0 to 5.0 mM) were prepared by mixing of appropriate volumes of the two stock solutions. The fluorous sensing phases (30 µL) were then applied with a micropipette onto a stack of 6 porous filter disks (porous poly(tetrafluoroethylene), without backing; 47 mm diameter, 0.45 m pore size, 50 m thick, 85% porosity, Fluoropore, Millipore) that mechanically supported the fluorous sensing phases, as described previously.34-35, 38-42 Full penetration of the sensing phases into the porous supports was spontaneous and was apparent from the translucence of the impregnated filter disks.

6

ACS Paragon Plus Environment

Page 7 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Electrode Assembly: The fluorous sensing membranes were mounted into custom-fabricated electrode bodies machined from poly(chlorotrifluoroethylene), as previously reported.35 In short, a cap with a center hole was screwed onto an electrode body, holding the sensing membrane between the cap and the electrode body but leaving a circular membrane area of 8.3 mm diameter exposed (see Figure 1 in ref. 35). An aqueous 0.10 mM AgNO3 solution was used as the inner filling solution, into which a AgCl-coated Ag wire was inserted as reference electrode. Before measurements, all electrodes were conditioned in a 10 mM AgNO3 solution for at least 10 h. EMF Measurements: Potentiometric measurements were performed at room temperature in stirred solutions with an EMF 16 potentiometer (Lawson Labs, Malvern, PA) controlled with EMF Suite 1.02 software (Fluorous Innovations, Arden Hills, MN). A free-flowing double-junction Ag/AgCl electrode with a 1.0 M LiOAc bridge electrolyte and AgCl-saturated 3.0 M KCl reference electrolyte was used as the external reference electrode (DX200, Mettler Toledo, Switzerland).43 Measurements were performed with polypropylene containers as sample beakers, which were cleaned weekly in 0.10 M HNO3. All emf values were corrected for liquid-junction potentials using the Henderson equation.44 Selectivity coefficients were determined with the separate solution method (SSM), fixed interference method (FIM), and fixed primary ion method (FPIM).45 In the concentration range where selectivities were measured, all ions responded Nernstian, as confirmed by successive dilutions of stock solutions. All reported selectivities are average values for 3 to 6 electrodes. This level of carefulness in the determination of selectivity coefficients is crucial because only unbiased selectivity coefficients45 are thermodynamically meaningful and lend themselves to the type of thorough interpretation as performed in this work.

RESULTS AND DISCUSSION Dependence of the Potentiometric Selectivity on the Ratio of Ionophore and Ionic Sites. The selectivity of an ISE membrane doped with an ionophore that forms only complexes of one type of 7

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

stoichiometry with the target ion (Figure 2A) exhibits a fairly simple dependence of the potentiometric selectivity on the concentration of the ionophore. This is illustrated by the dashed line in Figure 3 for an ionophore that forms 4:1 complexes with the target ion. For ionophore concentrations that are too low in comparison to the ionic site concentration, not all target ions in the ISE membrane can form complexes. As a result, the high concentration of unbound target ions in the membrane results in an ISE selectivity over non-complexing interfering ions that is very close to the selectivity of an ionophore-free ion-exchanger membrane (for a definition of the latter, see p 1595 of ref. 5). As the ratio of ionophore to ionic sites is increased, there is a rapid increase in selectivity as the total ionophore concentration exceeds the minimum amount necessary to bind all target ions in the membrane. At these high ionophore-to-ionic site ratios, the free ionophore concentration in the membrane is large, and the free target ion concentration in the membrane is lowered by many orders of magnitude below the complex concentration. This is the region where the free ionophore and the complex buffer the free target ion concentration in the membrane, which is a characteristic of a well-functioning ISE.

Figure 2. Schematic representing the effect of multiple complex stoichiometries on the composition of ISE membranes: (A) The ionophore, L, forms exclusively 4:1 complexes with the target ion, I+; the concentration of free I+, [I+], is very small if the complex formation constant is high and the concentration of R–, [R–], is smaller than one fourth of the total ionophore concentration, [Ltot]. (B) The ionophore forms complexes of 1:1, 2:1, 3:1, and 4:1 stoichiometry; the ratio of the concentrations of I+ 8

ACS Paragon Plus Environment

Page 9 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

and the different complexes depends on the formation constants of the individual complexes and on the ratio of [Ltot] and [R–].

+ Figure 3. Selectivities, 𝐾pot IJ , of ISE membranes for an ion I that forms multiple complexes with the

ionophore, L, with respect to an interfering ion J+ that does not bind to L. Shown are calculated values – –3 –10 M–1,  –10 M–2,  of log 𝐾pot IL2 = 10 IL3 = IJ versus [Ltot] for [R ] = 10 M, KIJex = 0.1, and (a) IL = 10

10–10 M–3, IL4 = 1048 M–4; (b) IL = 1018 M–1, IL2 = 1032 M–2, IL3 = 1042 M–3, IL4 = 1048 M–4.

While the above is well documented, it has not been discussed in detail in prior literature how the potentiometric selectivity depends on the total membrane concentration of an ionophore that forms complexes of multiple stoichiometries with the target ion. Intuitively, one might expect that the selectivity for the target ion over a non-complexing ion exhibits multiple steps when it is plotted against the ratio of the ionophore and ionic site concentrations, [Ltot]/[R]. As [Ltot]/[R] is increased, steps may be expected whenever [Ltot]/[R] exceeds a threshold value that allows for the exclusive formation of a ILnzi + n zL complex, where zi and zL are the charges of the target ion and the ionophore, respectively, and n is the number of ionophore molecules forming a complex ILnzi + n zL with ion i. At these specific [Ltot]/[R] ratios, the bulk of the sensing membrane contains only negligibly small

9

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

concentrations of free ionophore and complexes of other stoichiometry. A decrease in selectivity is expected at lower [Ltot]/[R] ratios because complexes of lower stoichiometries are formed due to the lack of available ionophore, and a higher selectivity is expected at higher [Ltot]/[R] ratios because either complexes of higher stoichiometries are formed or substantial concentrations of free ionophore are present. Taking into account both electrically neutral and charged ionophores, specific values of [Ltot]/[R] at which selectivity steps may be observed are | n zR / (zi + n zL) |, where zR is the charge of the ionic sites. For example, for a simple case of a monovalent anionic site, an electrically neutral ionophore, and a monovalent target cation that can bind up to 4 ionophore molecules, steps in the selectivity curve might be expected at [Ltot]/[R] values of 1, 2, 3, and 4. Such selectivity steps are indeed predicted for unique combinations of complex stability constants and concentrations of ionic sites, as illustrated by the solid line in Figure 3. (Notably, for a given set of complex stabilities, complexes of higher stoichiometries are favored by higher ionic site and ionophore concentrations, as illustrated by Figure S1 of the Supporting Information.) However, as shown in the following, the absence of multiple distinct selectivity steps should not be misinterpreted as an indication that only complexes of one stoichiometry are being formed. Moreover, even when distinct selectivity steps are missing, the careful interpretation of a selectivity plot as depicted in Figure 3 can still give crucial information about the types and stabilities of complexes that are formed. Calculation of Selectivity Coefficients and Membrane Compositions. Potentiometric selectivities and corresponding species concentrations in ISE membranes were calculated in this work for a complex-forming ion I+ and an interfering ion J+ that does not form complexes with L, as shown in Figures 3 to 10 and S1 to S4. To do so, the well-established phase boundary model4, 19 was used, taking into account all relevant equilibria in the ISE membranes. Briefly, considering a monovalent target cation and an electrically neutral ionophore, complex formation constants, ILn, were defined as 10

ACS Paragon Plus Environment

Page 11 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

ILn = [ILn+] / ([I+] [L]n)

(1)

For an ionophore that forms 1:1, 2:1, 3:1, and 4:1 complexes, the mass balance for the ionophore in the ISE membrane is given by [Ltot] = [L] + [IL+] + 2 [IL2+] + 3 [IL3+] + 4 [IL4+]

(2)

Assuming a monovalent anion as ionic site, R–, electroneutrality in the bulk of the ISE membrane requires that [R–] = [I+] + [IL+] + [IL2+] + [IL3+] + [IL4+]

(3)

The selectivity, 𝐾pot IJ , of the ISE membrane is given by (eqn 33 in ref. 4). + + 𝐾pot IJ = KIJex [I (I)] / [J (J)]

(4)

where [I+(I)] and [J+(J)] represent the concentrations of I+ and J+ on the ISE membrane side of the phase boundary layer between the ISE and the sample when I+ and J+ are measured separately, i.e., when the membrane contains either only I+ and ILn+ complexes or only J+ and JLn+ complexes. KIJex is the equilibrium constant that describes the ion exchange of uncomplexed target and interfering ions between the aqueous sample and the ISE membrane: KIJex = aI(aq) aJ(aq)–1 [J+] [I+]–1

(5)

where aI(aq) and aJ(aq) represent the activities of the respective ions in the aqueous phase. The set of eqns 2, 3, 4, and the 4 equations corresponding (for n = 1, 2, 3, and 4) to eqn 1 cannot be solved algebraically to give 𝐾pot IJ as a function of [Ltot]. However, the set of eqns 2, 3, and the four equations corresponding to eqn 1 can be readily solved to give [Ltot] as a function of [L]. Moreover, + 𝐾pot IJ can also be obtained as a function of [L] by (i) solving for [I ] the set of eqn 3 and the four

equations corresponding to eqn 1, (ii) replacing in eqn 4 the term [I+(I)] with the thus obtained expression for [I+], and (iii) replacing in the resulting equation the term [J+(J)] with [R–] (because J+ does not form a complex with L and, therefore, the two concentrations equal one another). So-called

11

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

parametric plots of 𝐾pot IJ versus [Ltot] can then be obtained by numerical calculation of corresponding 𝐾pot IJ and [Ltot] pairs for a range of [L] values. Parametric plots of the membrane concentrations of the various complexes in ISE membranes can be obtained similarly by (i) solving for [I+] the set of eqn 3 and the 4 equations corresponding (for n = 1, 2, 3, and 4) to eqn 1, and (ii) solving eqn 1 for [ILn+] for the value of n of interest, followed by insertion of [I+] from (i) into the resulting equation. This gives [ILn+] as a function of [L]. Parametric plot of [ILn+] versus [Ltot] are again obtained upon numerical calculation of corresponding [ILn+] and [Ltot] pairs for a range of [L] values. For further details see the Supporting Information. Multiple Complexes ILn+ with Equal Stepwise Binding Constants. Let us first consider a system in which all four complexes IL+, IL2+, IL3+, and IL4+ are formed, assuming that the stepwise binding constants KIL, KIL2, KIL3, and KIL4 for each of these complexes are the same, i.e., if the free energy of binding of L to ILn-1+ to give ILn+ is identical for all n: KILn = [ILn+] / ([ILn–1+] [L] ) = ILn / ILn-1

(6)

where ILn and ILn-1 represent cumulative binding constants (i.e.: ILn = [ILn+] [I+]–1 [L]–n). Figure 4 shows the predicted selectivity for four different systems with stepwise binding constants, KILn, of 103, 104, 105, and 106 M–1 for (a), (b), (c), and (d), respectively. As expected, the larger the complex stability is, the larger the predicted selectivity at high total ionophore concentrations. However, despite the formation of IL+, IL2+, and IL3+ complexes of substantial stability, none of the selectivity curves shows more than one step, very unlike the pattern shown in Figure 3 (solid line).

12

ACS Paragon Plus Environment

Page 13 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 4. Selectivities of ISE membranes for an ion I+ that forms multiple complexes with the ionophore, L, with respect to an interfering ion J+ that does not bind to L. Shown are calculated values – –3 3 –1 6 –2 9 of log 𝐾pot IJ versus [Ltot] for [R ] = 10 M, KIJex = 0.1 and (a) IL = 10 M , IL2 = 10 M , IL3 = 10

M–3, IL4 = 1012 M–4; (b) IL = 104 M–1, IL2 = 108 M–2, IL3 = 1012 M–3, IL4 = 1016 M–4; (c) IL = 105 M–1, IL2 = 1010 M–2, IL3 = 1015 M–3, IL4 = 1020 M–4; (d) IL = 106 M–1, IL2 = 1012 M–2, IL3 = 1018 M– 3,

IL4 = 1024 M–4.

For ionophore-to-ionic site ratios larger than 4, the selectivity predicted for a membrane in which IL+, IL2+, IL3+, and IL4+ can all be formed is indistinguishable from the selectivity for the membranes in which only IL4+ is formed. This is illustrated by Figure 5, which shows the solid line for pot + + + + log 𝐾pot IJ of membranes in which IL , IL2 , IL3 , and IL4 can be formed and the dashed line for log 𝐾IJ

of membranes in which only IL4+ can be formed. Figure 5 shows that below the ratio of total ionophore to ionic site of 4:1, despite the possibility for the formation of IL+, IL2+, and IL3+ complexes, the selectivity of the membrane in which IL+, IL2+, IL3+, and IL4+ can all be formed (solid line) is only slightly higher than for an ionophore-free ion-exchanger membrane (dotted line) and the membrane in which only IL4+ can be formed (dashed line).

13

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

Figure 5. Selectivities of ISE membranes for an ion I+ that forms multiple complexes with the ionophore, L, with respect to an interfering ion J+ that does not bind to L. Shown are calculated values – –3 6 –1 12 –2 18 of log 𝐾pot IJ versus [Ltot] for [R ] = 10 M, KIJex = 0.1 and (a) IL = 10 M , IL2 = 10 M , IL3 = 10

M–3, IL4 = 1024 M–4; (b) IL = 10–10 M–1, IL2 = 10–10 M–2, IL3 = 10–10 M–3, IL4 = 1024 M–4. (c) For comparison, selectivity of an ionophore-free ion-exchanger electrode.

+ Plots of log 𝐾pot IJ versus [Ltot] in systems in which ILn complexes are formed with identical

stepwise complex formation constants, KILn, also show an absence of multiple steps when the maximum number of ionophore molecules allowed to bind to I+ is varied between 1 and 4 (see Figure S2 of the Supporting Information for systems with KILn = 106 M–1). Stepwise Binding Constants Decreasing in Strength with Stoichiometry. Comparison of Figures 3, 4, and S2 suggests that steps in the plot of the selectivity versus the ionophore-to-ionic site ratio are only possible if the stepwise binding constants decrease with increasing complex stoichiometry. Even if that is true, a selectivity step is not always observed though. This is illustrated in Figure 6 for a system in which only 1:1 and 2:1 complexes can be formed. For all five sets of stability constants, IL has the same value of 109 M–1. For KIL2 = 10–10 M–1 (a) and KIL2 = 104 M–1 (b), the 1:1 complex dominates how the selectivity depends on the ionophore concentration, resulting in a rapid 14

ACS Paragon Plus Environment

Page 15 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

increase in selectivity only at an ionophore-to-ionic site ratio of 1:1. The opposite is true for KIL2 = 108 M–1 (d) and KIL2 = 1010 M–1 (e), for which a rapid increase in selectivity is seen only at an ionophoreto-ionic site ratio of 2:1. Two separate distinctive jumps in selectivity at the ionophore-to-ionic site ratios 1:1 and 2:1 are observed only for KIL2 = 106 M–1 (c). This is consistent with the conclusion that multiple selectivity steps are only possible if the stepwise binding constants decrease with increasing complex stoichiometry. Interestingly, the concentrations of the membrane species corresponding to Figure 6 (shown in Figure S3 of the Supporting Information) show that at the ionophore-to-ionic site ratio of 1:1 there is a substantial concentration of IL+ for all of the KIL2 values considered (99.4, 94.0, 61.2 and 13.7% of the total ionophore concentration for the values of KIL2 of 104, 106, 108, and 1010 M–1, respectively). Even when KIL2 exceeds KIL by one order of magnitude and the selectivity appears to be clearly indicative of 2:1 complexation (i.e., curve (e)), the ratio of [IL2+] to [IL+] at the total ionophore-to-ionic site ratio of 1:1 is still only 43.2 : 13.7. This example illustrates how easily multiple complex stoichiometries can remain unrecognized in a qualitative interpretation of log KILn versus [Ltot] relationships.

Figure 6. Selectivities of ISE membranes for an ion I+ that forms IL+ and IL2+ complexes with the ionophore, L, with respect to an interfering ion J+ that does not bind to L. Shown are calculated values

15

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

– –3 9 –1 –10 M–2, (b)  of log 𝐾pot IL2 = IJ versus [Ltot] for [R ] = 10 M, KIJex = 0.1, IL = 10 M , and (a) IL2 = 10

1013 M–2, (c) IL2 = 1015 M–2, (d) IL2 = 1017 M–2, and (e) IL2 = 1019 M–2.

The Example of [9]Mercuracarborand-3 ISEs. An experimental example that illustrates how easy it can be to overlook ionophores that form complexes of multiple stoichiometries with the primary ion is given by [9]mercuracarborand-3. An initial report on ISEs based on this ionophore reported 1:1 complex formation with Cl-.46 A later study of optode membranes doped with this ionophore found inconsistencies and clarified that this ionophore forms 1:1 and 2:1 complexes with Cl–, with overall complex formation constants of IL = 109.9 M–1 and IL2 = 1013.4 M–2, respectively.

21-22

Using these

values of IL and IL2 and adapting the model described above for the [9]mercuracarborand-3 system, the dependence of log 𝐾pot ClJ on the total ionophore concentration as shown by curve (b) in Figure 7 is predicted. For comparison, curves (a) and (c) show log 𝐾pot ClJ as calculated for hypothetical ionophores that form either only 1:1 or only 2:1 complexes, respectively, using the same complex formation constants as for (b). Qualitatively, the most noticeable feature of curve (b) is the abrupt change in selectivity at the ionophore-to-ionic site ratio of 1:1. A second large step in selectivity is missing. This could easily be misinterpreted as an indication for unique 1:1 complex stoichiometry. However, as the comparison of curves (b) and (a) shows, log 𝐾pot ClJ is noticeably larger for ionophore-to-ionic site ratios between 1 and 2 than for a system with unique 1:1 stoichiometry. Moreover, at ionophore-to-ionic site ratios larger than 2, log 𝐾pot ClJ is noticeably smaller than for unique 1:1 stoichiometry. Quite remarkable is also the substantial increase in selectivity (i.e., decrease in log 𝐾pot ClJ ) in the range from ionophore-toionic site ratios of 1 to 2, as compared to a system in which only 2:1 complexes are formed (i.e., curve (c)), despite the fact that the complexation constant KClL2 is with 103.5 M–1 relatively small in comparison to the KClL of 109.9 M–1. The comparatively weak stability of the 2:1 complex is reflected by the rather modest decrease in log 𝐾pot ClJ from curve (a) to curve (c) at high total ionophore 16

ACS Paragon Plus Environment

Page 17 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

concentrations. Indeed, at an ionophore-to-ionic site ratio of 2:1, 11% of the ionophore is in its free form, while at the ionophore-to-ionic site ratio of 1:1 only 0.06% of the ionophore is in its free form, which is consistent with a relatively low affinity of ClL– for a second ionophore ligand. Arguably, the reported conclusion that the 2:1 complex is “extremely stable” is correct in the sense that ClL2– formation from Cl– and two ionophore molecules can result in Donnan failure. However, with a view to formation of ClL2– from ClL– and L, the ClL2– complex has only a comparatively low stability.22 (Figure S4 of the Supporting Information depicts for the [9]mercuracarborand-3 system the concentration of all membrane species as a function of the ratio of the ionophore and ionic sites concentrations.)

Figure 7. Selectivities of ISE membranes for an ion I– that forms IL– and IL2– complexes with the ionophore, L, with respect to an interfering ion J– that does not bind to L. Shown are calculated values 11.3 M–1,  –10 + –3 of log 𝐾pot IL2 = 10 IJ versus [Ltot]/[ionic sites] for [R ] = 510 M, KIJex = 1, and (a) IL = 10

M–2; (b) IL = 109.9 M–1, IL2 = 1013.4 M–2; (c) IL = 10–10 M–1, IL2 = 1013.4 M–2. Inset: Structure formula of [9]mercuracarborand-3.

Site Optimization in Case of Primary and Interfering Ions That Form Complexes of Multiple Stoichiometries. One of the most important tasks in developing a new ISE is the choice of

17

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 29

the optimum ratio of ionophore and ionic sites, so that the highest possible potentiometric selectivity is achieved. While this process was entirely empirical in the early days of ISE development, guidelines for optimum ionophore-to-ionic site ratios were introduced first for electrically neutral ionophores14-16 and subsequently for electrically charged ionophores.17-18 Even when complex stoichiometries are not known prior to potentiometric measurements, those guidelines are extremely useful because, based on educated guesses of probable stoichiometries, they provide for each system of interest a small number of ionophore-to-ionic site ratios that are most likely to provide the highest selectivities. Unfortunately, those guidelines no longer apply the same way if the primary ion, interfering ion, or both form complexes of more than one stoichiometry. An example that illustrates this is given by ISE membranes doped with an ionophore that binds a monovalent primary cation, I+, with 1:1 stoichiometry and a monovalent interfering ion, J+, with both 1:1 and 2:1 stoichiometry. Curve (a) in Figure 8 shows the predicted log 𝐾pot IJ as a function of the total ionophore concentration for the case of extremely weak 1:1 but very stable 2:1 complexes of J+ (IL2 = 1015 M–2). Established theory predicts the highest selectivity for this system for 71 mol % ionic sites with respect to the ionophore,18 which in Figure 8 (calculated for an anionic site concentration of 1.0 mM) corresponds to a total ionophore concentration of 1.41 mM. Allowing also for 1:1 complex formation of J+ with complex stabilities of JL of 107, 108, 109, and 1010 M–1 not only gradually worsens the optimum selectivity but it also shifts the selectivity optimum steadily closer and closer to a total ionophore concentration of 1.0 mM (curves (b) to (e)). Recognizing this shift in the optimum concentration of ionic sites is important for the optimization of the ISE’s selectivity. For example, the inadequate assumption of exclusive 2:1 complex formation for J+ and ensuing choice of 71 mol % ionic 10 –1 sites based on traditional ionic site theory would give a log 𝐾pot IJ of –0.56 for JL = 10 M (curve (e)),

whereas the optimum site ratio of 98% is predicted to give a 𝐾pot IJ of –0.98.

18

ACS Paragon Plus Environment

Page 19 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 8. Selectivities of ISE membranes for a primary ion I+ that forms only IL+ complexes (IL = 109 M–1) with the ionophore, L, with respect to an interfering ion J+ that that forms both JL+ and JL2+ – – complexes with L (IL2 = 1015 M–2). Shown are calculated values of log 𝐾pot IJ versus [Ltot] for [R ] = 10 3

M, KIJex = 1, and (a) JL = 10–10 M–1, (b) JL = 107 M–1, (c) JL = 108 M–1, (d) JL = 109 M–1, and (e)

JL = 1010 M–1.

Another notable fact illustrated by curve (e) is that ISE membranes with an optimized ionophore-to-site ratio may be selective for I+ over ion J+ if the ionophore forms a 2:1 complex only with J+, even when the 1:1 complex of the primary ion, I+, is less stable than the 1:1 complex with the interfering ion, J+. It is interesting to consider the reasons for this selectivity at the molecular level. In the case of an ISE membrane with the optimized 71 mol % anionic site concentration and an ionophore that forms only 1:1 complexes with I+, exposure of this membrane to aqueous solutions of I+ results in 71% of the ionophore in the membrane bulk forming complexes with I+, and 29% of the ionophore remaining in the uncomplexed form. If the same membrane is exposed to samples that contain an ion J+ that forms only 2:1 complexes, all the ionophore is in the form of the complex JL2+, and the membrane contains an additional 21 mol % uncomplexed J+. The excess of ionophore in the case of I+ exposure combined with the excess of uncomplexed J+ in the case of J+ exposure is the cause for the high

19

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

selectivity for I+ over J+. Instead, if the ion J+ forms both 2:1 and 1:1 complexes, exposure of this membrane with 71% ionic sites to aqueous solutions of J+ results in 58% of the ionophore forming 2:1 complexes with J+ and 42% of the ionophore forming 1:1 complexes with J+. The good selectivity for I+ over J+ is the result of an excess of ionophore in the case of I+ exposure, combined with the extremely low concentration of free ionophore in the case of J+ exposure (but not a large concentration of uncomplexed J+ in the bulk of the ISE membrane). Because it appears improbable that an ionophore that forms 2:1 complexes with an ion will not form 1:1 complexes at all, the ISE literature likely contains a substantial number of publications that failed to recognize complexes of multiple stoichiometries for a given ion. Fluorophilic Ag+ Ionophore 2. A detailed study of potentiometric selectivities over a large range of ionophore-to-ionic site ratios has not been performed for any of the above-mentioned systems in which a primary ion forms complexes of multiple stoichiometries. To this end, we reinvestigated the properties of the fluorophilic Ag+ ionophore 2. We previously showed that ISEs with fluorous membranes doped with 2 exhibit selectivities of nine to eleven orders of magnitude over K+, Na+, and Cu2+.40, 47-48 For this work, we focused on the selectivities for Ag+ over the tetraethylammonium ion, Et4N+, because we wanted to avoid complications from multiple stoichiometries for both the primary and the interfering ion. While tetraalkylammonium ions have been shown to bind in fluorous solvents to trialkylamine derivatives by formation of hydrogen bonds of the type R3N+–C(R)–H…NR3 and should not be considered inert under all circumstances,49 binding of Et4N+ to thioethers is not a concern. Thioethers have a pKa in water of approximately –5.4 (value estimated for (CH3)2SH+),50 a basicity that is approximately 15 orders of magnitude lower than for trialkylamines. Therefore, Et4N+ was assumed for this work to be an ion that does not bind to the ionophore. Figure 9 shows log 𝐾pot Ag,Et4N for ionophore-to-ionic site concentration ratios from 0 to 4. The experimental data were fitted with a model taking into account AgL+ and AgL2+ complexes. As curve (a) shows, the fit is very good, suggesting that the bidentate ionophore 2 forms a tetracoordinated Ag+ 20

ACS Paragon Plus Environment

Page 21 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

complex and that there is no need for consideration of complexes of higher stoichiometry. This is in agreement with the literature, which reports a strong preference of Ag+ for low coordination numbers (see ref. 40 and references cited therein).

Figure 9. ISE membrane consisting of perfluoroperhydrophenanthrene doped with ionic sites 1 and ionophore 2: Potentiometric selectivities for Ag+ over Et4N+ as a function of the ratio of ionophore and ionic sites ([R+] = 10–3 M); determined with the separate solution method (see Table S1 in the Supporting Information for numerical values). (a) Experimental log 𝐾pot Ag,Et4N values were fitted with the model described in the text, giving the following parameters: KAg,Et4N,ex = 2.6x103, AgL = 7.9  104 mol–1 kg–1, AgL2 = 1.2  108 mol–2 kg–2. Also shown are selectivities as predicted for the same stability constants but formation of (b) 1:1 complexes only (i.e., AgL = 7.9  104 mol–1 kg–1, AgL2 = 10–10 mol–2 kg–2 ) or (c) 2:1 complexes only (i.e., AgL = 10–10 mol–1 kg–1, AgL2 = 1.2  108 mol–2 kg–2).

The fitted values of the stepwise complex formation constants KAgL and KAgL2 of 7.9  104 and 1.5  103 mol–1 kg-1, respectively, are consistent with a common observation for organometallic complexes, i.e., that KIL > KIL2. Note that the value of KAgL2 is small enough to result in a significant increase in selectivity near the ionophore-to-ionic site ratio of 1:1, which would not be expected in the hypothetical case of exclusive 2:1 complex formation with the same AgL2 (curve c). However, KAgL2 is 21

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 29

also large enough to cause a substantial increase in selectivity at high ionophore concentrations, as compared to the hypothetical case of exclusive 1:1 complex formation with the same value of AgL, as illustrated in Figure 9 by curve (b). As shown in Figure 10, in the range of ionophore-to-ionic site ratios from 1 to 5, the cause for the gradual but substantial increase in selectivity is the increase in free ionophore concentration, which steadily changes the ratio of the AgL+ and AgL2+ complex concentrations from 0.10 to 0.84 and 4.77 at the ionophore-to-ionic site ratios of 1:1, 2:1, and 5:1, respectively.

Figure 10. Concentrations of the free ionophore 2 and its complexes with Ag+ as calculated from the fit of log 𝐾pot Ag,Et4N shown in Figure 9.

CONCLUSIONS With few exceptions, the ISE literature has overlooked in the past the concurrent formation of primary ion complexes with multiple stoichiometries. This is surprising because typically chemical interactions that lead to the formation of ILn+ complexes will also stabilize complexes of a lesser stoichiometry. Because complexation equilibria with KILn >> KILn-1 are few, exclusive formation of complexes such as IL2+ or IL3+ is unlikely unless the ionophore is used in excess (as illustrated in Figures 6 and S2).

22

ACS Paragon Plus Environment

Page 23 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

The failure to notice the possible formation of ILn-1+ complexes of the primary ion is of lesser concern if the ratio of ionophore and ionic site is optimized in view of the formation of ILn+ complexes. However, if KILn is not sufficiently large, ILn+ will coexist with ILn–1+. In such a case, a higher potentiometric selectivity will be observed at ionophore-to-ionic site ratios that are larger than what one might expect from the n:1 stoichiometry alone (as shown for [9]mercuracarborand-3 in Figure 7 and for Ag+ ionophore 2 in Figures 9 and 10). Formation of JLn-1+ complexes of the interfering ion cannot be ignored either as that can affect the ionophore-to-ionic site ratio that provides the optimum selectivity (as illustrated in Figure 8). To assure that ISE membranes are formulated to provide the highest possible selectivities with a given ionophore, experimental selectivities should be determined for a range of ionophore-to-ionic site pot ratios. Plots of log 𝐾pot IJ versus [Ltot] are unlikely to exhibit multiple abrupt changes in log 𝐾IJ that

indicate different complex stoichiometries (as for the hypothetical example of Figure 3), but fits of experimental selectivities can provide both stoichiometries and complexation constants (as shown, e.g., in Figure 9). Alternatively, potentiometric selectivities for systems with multiple complex equilibria may be predicted mathematically using complex formation constants determined with other techniques, such as the sandwich membrane technique, optical methods (such as for Figure 7), or ion transfer voltammetry.51-54 Whichever experimental approach is taken, investigators should be prepared to consider the concurrent presence of multiple complexes for a given ion, be it the primary ion or interfering ions. Failure to do so can result not only in the preparation of ISE membranes with inferior selectivity, but it may also provide misleading guidance for the design of new and better ionophores.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. 23

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 29

pot Log 𝐾pot IJ in systems analogous to Figure 3 but with different ionic site concentrations. Log 𝐾IJ

in systems in which ILn+ complexes are formed with identical stepwise complex formation constants. Species concentrations in ISE membranes corresponding to Figures 6 and 7. Numerical values of 𝐾pot Ag,Et4N corresponding to Figure 9. Details on the preparation of parametric plots. AUTHOR INFORMATION Corresponding Author * E-mail: [email protected]. Tel: +1 (612) 624-1431. ORCID Ibrahim Yilmaz: 0000-0002-9447-3065 Rosenildo Correa da Costa: 0000-0002-1005-8455 John A. Gladysz: 0000-0002-7012-4872 Philippe Bühlmann: 0000-0001-9302-4674 Evan Anderson: 0000-0002-2506-1147

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENTS I.Y. would like to thank TUBITAK for a fellowship throughout this study. This work was supported by the National Science Foundation (CHE-1748148) and the Welch Foundation (Grant A-1656). E.L.A. thankfully acknowledges a Lester C. and Joan M. Krogh Endowed Fellowship, an ACS Division of Analytical Chemistry and Eastman Summer Fellowship, and a Dissertation Fellowship from the Graduate School, University of Minnesota. 24

ACS Paragon Plus Environment

Page 25 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

REFERENCES 1.

Amemiya, S., Potentiometric Ion-Selective Electrodes. In Handbook of Electrochemistry, Zoski,

C. G., Ed. Elsevier: Amsterdam, 2007. 2.

De Marco, R.; Clarke, G.; Pejcic, B., Electroanalysis 2007, 19, 1987-2001.

3.

Bobacka, J.; Ivaska, A.; Lewenstam, A., Chem. Rev. 2008, 108, 329-351.

4.

Bakker, E.; Bühlmann, P.; Pretsch, E., Chem. Rev. 1997, 97, 3083-3132.

5.

Bühlmann, P.; Pretsch, E.; Bakker, E., Chem. Rev. 1998, 98, 1593-1687.

6.

Bakker, E.; Pretsch, E., Angew. Chem., Int. Ed. 2007, 46, 5660-5668.

7.

Bühlmann, P.; Chen, L. D., Ion-Selective Electrodes With Ionophore-Doped Sensing Membranes.

In Supramolecular Chemistry: From Molecules to Nanomaterials; Volume 5: Self-Assembly and Supramolecular Devices, John Wiley & Sons: Chichester, UK, 2012; pp 2539–2580. 8.

Yin, T. J.; Qin, W., TrAC, Trends Anal. Chem. 2013, 51, 79-86.

9.

Van den Berg, A.; Van der Wal, P. D.; Skowronska-Ptasinska, M.; Sudholter, E. J. R.; Reinhoudt,

D. N., Anal. Chem. 1987, 59, 2827-2829. 10. Buck, R. P.; Toth, K.; Graf, E.; Horvai, G.; Pungor, E., J. Electroanal. Chem. 1987, 223, 51-66. 11. Bühlmann, P.; Yajima, S.; Tohda, K.; Umezawa, Y., Electrochim. Acta 1995, 40, 3021-3027. 12. Bühlmann, P.; Yajima, S.; Tohda, K.; Umezawa, K.; Nishizawa, S.; Umezawa, Y., Electroanalysis 1995, 7, 811-816. 13. Gyurcsanyi, R. E.; Lindner, E., Anal. Chem. 2002, 74, 4060-4068. 14. Meier, P. C.; Morf, W. E.; Läubli, M.; Simon, W., Anal. Chim. Acta 1984, 156, 1-8. 15. Ammann, D.; Pretsch, E.; Simon, W.; Lindner, E.; Bezegh, A.; Pungor, E., Anal. Chim. Acta 1985, 171, 119-129. 16. Eugster, R.; Gehrig, P. M.; Morf, W. E.; Spichiger, U. E.; Simon, W., Anal. Chem. 1991, 63, 2285-2289. 25

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 29

17. Schaller, U.; Bakker, E.; Spichiger, U. E.; Pretsch, E., Anal. Chem. 1994, 66, 391-398. 18. Amemiya, S.; Bühlmann, P.; Pretsch, E.; Rusterholz, B.; Umezawa, Y., Anal. Chem. 2000, 72, 1618-1631. 19. Bakker, E.; Bühlmann, P.; Pretsch, E., Talanta 2004, 63, 3-20. 20. Eugster, R.; Spichiger, U. E.; Simon, W., Anal. Chem. 1993, 65, 689-695. 21. Malon, A.; Radu, A.; Qin, W.; Qin, Y.; Ceresa, A.; Maj-Zurawska, M.; Bakker, E.; Pretsch, E., Anal. Chem. 2003, 75, 3865-3871. 22. Ceresa, A.; Qin, Y.; Peper, S.; Bakker, E., Anal. Chem. 2003, 75, 133-140. 23. Miyake, M.; Chen, L. D.; Pozzi, G.; Bühlmann, P., Anal. Chem. 2012, 84, 1104-1111. 24. Horvath, I. T.; Rabai, J., Science 1994, 266, 72-75. 25. Gladysz, J. A.; Curran, D. P.; Horváth, I. T., Handbook of Fluorous Chemistry. Wiley/VCH: Weinheim, 2004. 26. Amemiya, S.; Bühlmann, P.; Umezawa, Y., Anal. Chem. 1998, 70, 445-454. 27. Bühlmann, P.; Umezawa, Y., Electroanalysis 1999, 11, 687-693. 28. Amemiya, S.; Bühlmann, P.; Odashima, K., Anal. Chem. 2003, 75, 3329-3339. 29. Kakiuchi, T.; Senda, M., J. Electroanal. Chem. 1991, 300, 431-445. 30. Kakiuchi, T.; Senda, M., Anal. Sci. 1991, 7, 591-594. 31. Kakiuchi, T., J. Electroanal. Chem. 1993, 345, 191-203. 32. Amemiya, S., Anal. Chem. 2016, 88, 8893-8901. 33. Greenawalt, P. J.; Garada, M. B.; Amemiya, S., Anal. Chem. 2015, 87, 8564-8572. 34. Boswell, P. G.; Bühlmann, P., J. Am. Chem. Soc. 2005, 127, 8958-8959. 35. Boswell, P. G.; Lugert, E. C.; Rabai, J.; Amin, E. A.; Bühlmann, P., J. Am. Chem. Soc. 2005, 127, 16976-16984. 36. Rocaboy, C.; Gladysz, J. A., Tetrahedron 2002, 58, 4007-4014. 37. da Costa, R. C.; Jurisch, M.; Gladysz, J. A., Inorg. Chim. Acta 2008, 361, 3205-3214. 26

ACS Paragon Plus Environment

Page 27 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

38. Chen, L. D.; Lai, C. Z.; Granda, L. P.; Fierke, M. A.; Mandal, D.; Stein, A.; Gladysz, J. A.; Bühlmann, P., Anal. Chem. 2013, 85, 7471-7477. 39. Chen, L. D.; Mandal, D.; Pozzi, G.; Gladysz, J. A.; Bühlmann, P., J. Am. Chem. Soc. 2011, 133, 20869-20877. 40. Lai, C. Z.; Fierke, M. A.; da Costa, R. C.; Gladysz, J. A.; Stein, A.; Bühlmann, P., Anal. Chem. 2010, 82, 7634-7640. 41. Chen, L. D.; Mandal, D.; Gladysz, J. A.; Bühlmann, P., New J. Chem. 2010, 34, 1867-1874. 42. Boswell, P. G.; Szijjarto, C.; Jurisch, M.; Gladysz, J. A.; Rabai, J.; Bühlmann, P., Anal. Chem. 2008, 80, 2084-2090. 43. Dohner, R.; Wegmann, D.; Morf, W. E.; Simon, W., Anal. Chem. 1986, 58, 2585-2589. 44. Morf, W. E., The Principles of Ion-Selective Electrodes and of Membrane Transport. Elsevier: New York, 1981. 45. Bakker, E.; Pretsch, E.; Bühlmann, P., Anal. Chem. 2000, 72, 1127-1133. 46. Badr, I. H. A.; Diaz, M.; Hawthorne, M. F.; Bachas, L. G., Anal. Chem. 1999, 71, 1371-1377. 47. Mousavi, M. P. S.; Gunsolus, I. L.; Perez De Jesus, C. E.; Lancaster, M.; Hussein, K.; Haynes, C. L.; Bühlmann, P., Sci. Total Environ. 2015, 537, 453-461. 48. Gunsolus, I. L.; Mousavi, M. P. S.; Hussein, K.; Bühlmann, P.; Haynes, C. L., Environ. Sci. Technol. 2015, 49, 8078-8086. 49. Anderson, E. L.; Gingery, N. M.; Boswell, P. G.; Chen, X. V.; Rabai, J.; Bühlmann, P., J. Phys. Chem. B 2016, 120, 11239-11246. 50. Reich, H. J. https://http://www.chem.wisc.edu/areas/reich/pkatable/ (accessed 5 August 2018). 51. Bakker, E.; Willer, M.; Lerchi, M.; Seiler, K.; Pretsch, E., Anal. Chem. 1994, 66, 516. 52. Ceresa, A.; Pretsch, E., Anal. Chim. Acta 1999, 395, 41-52. 53. Mi, Y. M.; Bakker, E., Anal. Chem. 1999, 71, 5279-5287.

27

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 29

54. Shultz, M. M.; Stefanova, O. K.; Mokrov, S. S.; Mikhelson, K. N., Anal. Chem. 2002, 74, 510517.

28

ACS Paragon Plus Environment

Page 29 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Table of Contents Image

29

ACS Paragon Plus Environment